Women on corporate boards around the world: Triggers and
barriers
Amon Chizema
University of Birmingham
Birmingham Business School
B15 2TY, Edgbaston, Birmingham, UK
&
Dzidziso Samuel Kamuriwo
Cass Business School
City University London,
106 Bunhill Row, EC1Y 8TZ, UK
&
Yoshikatsu Shinozawa
SOAS, University of London
London WC1H 0XG, London, UK
1
Women on corporate boards around the world: Triggers and
barriers
Abstract
One of the institutions in which the gender gap remains a contestable issue is the board
of directors, where the proportion of female directors is still low. While some countries
have achieved higher proportions of female directors on their corporate boards, others
have not registered even a single one. Drawing on social role theory, that places
emphasis on traditional gender activities, this study starts by arguing that board
directorship is an agentic role and more suitable for men. The study shows that key
social institutions have the potential to alleviate such stereotypical attitudes or to
maintain the status quo. Employing a robust statistical technique in two-stage least
squares (2SLS), this study finds that the representation of women in other key national
institutions, such as in politics, positively affects the appointment of female directors
on boards. On the other hand, religiosity has a negative causal effect on female board
appointments.
Key words: female directors, board of directors, social role theory, social institutions
2
Introduction
The lack of female representation on corporate boards has attracted the attention
of academics (Hillman, Shropshire & Cannella, 2007), policy makers (OECD, 2009),
practitioners and civil society (Catalyst, 2007). Part of this interest derives from the
observed and perceived importance of achieving gender equity in societal and political
leadership (Bullough, Kroeck, Newbury, Kundu & Lowe, 2012) as well as in firms
(Nielsen & Huse, 2010).
Explanations for the representation of women in elite leadership roles
traditionally focused on the idea that a lack of qualified women created a “pipeline
problem”. This shortage of qualified women has been ascribed to a variety of causes,
including women’s family responsibilities (Greenhaus & Parasuraman, 1999) and
inherited tendencies for women to display fewer of the traits and motivations that are
necessary to attain and achieve success in high-level positions (Browne, 1999; Carter,
D'Souza, Simkins & Simpson, 2010; Goldberg, 1993).
Another explanation resides in gender role attitudes, “the opinions and beliefs
about the ways that family and work roles do and should differ based on sex” (Harris
& Firestone, 1998: 239). Indeed, some societies hold traditional gender role attitudes
and believe in a clear division of labor, where “men must be more concerned with
economic and other achievements, while women must be concerned with taking care
of people in general and of children in particular” (Hofstede, 2001: 280). In contrast,
non-traditional gender role attitudes in some societies suggest a less distinct genderrole-based division of labor, where men and women share various responsibilities (Van
Yperen & Buunk, 1991). We argue that, this view, consistent with insights from social
psychology on gender differences and social roles, (Wood & Eagly, 2012) may provide
better understanding of gender disparity on corporate boards.
3
Social role theory potentially undergirds the lack of female representation,
positing that differences in societal roles lead men and women to demonstrate and value
different types of interpersonal behaviors (Eagly, 1987; Eagly & Johannesen-Schmidt,
2001). The perception is that men tend to value and engage in more assertive,
competitive and agentic behaviors, whereas, because women traditionally occupy more
caretaking roles, they tend to value and engage in more communal behaviors (Eagly,
1987; Koenig, Eagly, Mitchell & Ristikari, 2011).This common placement of women
and men into prescribed roles produces gender stereotypes by observing their behavior
in such roles. Such stereotypical perceptions are shared by individuals in a society, from
where they derive their legitimacy, and are thus culturally consensual.
Applying the foregoing reasoning, in the context of the workplace and indeed
the boardroom, women are, therefore, more likely than men to hold positions at low
levels in hierarchies of status and authority and are less likely to be at the highest levels
of organizational hierarchies (Heilman, 2001; Wood & Eagly, 2012) where agency is
expected.
Notwithstanding the power and commonality of gender role perceptions, there
is evidence that in some hunter-gatherer societies, even men sometimes perform
substantial infant care (Fouts, 2008). Moreover, in many industrialized societies some
men pursue female-dominated professions such as nursing or social work (Sayer,
Cohen & Casper, 2004). The fact that men and women sometimes engage in gender
atypical activities suggests flexible behavior that is not rigidly differentiated by sex
(Wood & Eagly, 2012), but one that is shaped by situational/contextual demands. This
implies that both sexes can be socially sensitive or aggressive, given appropriate
socialization and support from the social environment, where beliefs and practices are
4
shared and subsequently modified (Richerson & Boyd, 2005; Tennie, Call &
Tomasello, 2009).
In addition, the fact that gender roles are a social construction, manifest in
stereotypes (Eagly & Karau, 2002), suggests that these [stereotypes] could be broken
down only by genuine social change. Such social change could see occupational and
domestic-work segregation weakened, consequently changing the perception of women
and men roles with an inclination to equality. Arguing for the inevitable rise of women’s
status, Jackson (1998) emphasizes that changes in economic, cultural and political
systems over the years alleviated women’s disadvantages by improving their access to
positions. As economies got better, production and power once the preserve of the
household (where men played a superior role) moved to economic and political
organizations, thus reducing men’s dominance. This observation underscores the
salience and role of social institutions in facilitating or inhibiting gender roles.
However, countries’ economic, political and cultural institutions have
progressed at differing rates. As such, at the country level, institutional environments,
and by extension, social institutions vary, suggesting different levels in the potential to
either minimize or enhance gender roles and stereotypes. Indeed, Jackson (1998) posits
that the timing, rate and form of specific changes around issues of gender inequality
have varied considerably across countries.
In the context of this study, we argue that some countries would, therefore, have
higher proportions of female directors on boards than others. Drawing on social role
theory, and in particular considering the extent to which gender role attitudes differ
among societies, this study seeks to explore the institutional factors that potentially lead
to varying levels of female board appointments across countries.
5
Two streams of literatures are of relevance to our study. The first stream relates
to the association between gender role attitudes and critical issues related to the
workplace environment for women (Eyring & Stead, 1998; Kirchmeyer, 2002). The
second examines the presence of women on boards within particular country settings
including the UK (Conyon & Mallin, 1997; Singh & Vinnicombe, 2004), the US
(Peterson & Philpot, 2007) Switzerland (Ruigrok, Peck &Tacheva, 2007) and Norway
(Nielsen & Huse, 2010).
While much has been written about this subject in single country studies, only
a few researchers (Grosvold & Brammer, 2011; Terjesen & Singh, 2008) have focused
their attention on cross-national studies. Using data from 38 countries over the years
2001 to 2007, Grosvold, & Brammer (2011) place countries in clusters defined by
national institutional systems, akin to varieties of capitalism (Hall & Soskice, 2001;
Jackson & Deeg, 2008), to understand the proportion of female directors on boards. On
their part, Terjesen & Singh (2008) find that countries with higher representation of
women on boards are more likely to have women in senior management and equal ratios
of male to female pay.
Our paper offers several contributions. First, this study employs a social
psychology lens i.e. social role theory to understand a topical corporate governance
issue on board gender diversity, not only in a single country but across forty-five
countries. Indeed, this work builds on previous studies, contributing to the growing
literature on gender diversity and the appointment of women to upper echelons of firms
across the world. As such, this study emphasizes the salience and diversity of the
institutional environments obtaining around the world (Aguilera & Jackson, 2003).
Second, our study is interdisciplinary, bringing together various strands of
literature: international business, corporate governance, gender, politics and
6
development studies. Such an approach facilitates better understanding of the subject
at hand, providing more awareness of the developments in various disciplines. Third,
this study improves our understanding of the macro-factors that determine gender
diversity of corporate boards. Fourth, this paper potentially helps companies to
understand the institutional environment in which they operate and more importantly
how to react to it when making board appointment decisions.
Theory and hypotheses
According to social role theory, the perceived differences in the behavior of
women and men originate in the contrasting distributions of men and women into social
roles (Eagly, 1987). Thus, men and women are thought to possess attributes that equip
them for sex-typical roles. For example, men are more likely than women to be
employed, especially in authority positions, and women are more likely than men to fill
caretaking roles at home as well as in employment settings (Eagly & Wood, 2012). As
such, gender differences and similarities in behavior reflect consensually-shared gender
role beliefs or gender stereotypes that in turn represent society’s perceptions of men’s
and women’s social roles in the community in which they live.
Sharing these beliefs as people in a society, gender roles get established and
accepted by individuals who constitute the community. Individually, people act on their
beliefs, recognize that others think similarly, and know that others act on this shared
knowledge (Ridgeway, 2006). As such, these typical attributes tend to be seen as
desirable and admirable for each sex, adding prescriptiveness to and consensus on
gender roles. Because these social roles are consensual, societies equip men and women
by undertaking extensive socialization to promote personality traits and skills that
facilitate role performance.
7
Thus, drawing mainly on social role theory, extant research suggests that men
and women respond differently to various aspects of social relationships, which can be
categorized as either communal or agentic (Eagly, 1987; Koenig, et al., 2011). The
communal dimension is interpersonally oriented and broadly described as a concern for
the welfare of others (e.g., nurturing, sympathetic, friendly), with women scoring higher
on this dimension than men (Eagly, 1987, 2009; Spence & Buckner, 2000). While these
stereotypes may be perceived as positive in content, however, they serve to justify
women’s continued acceptance of their traditional social roles (Glick & Fiske, 2001).
Moreover, they can undoubtedly possess negative consequences for women, as they are
suggested to undermine perceptions of competence and power (Jost & Kay, 2005).
On the other hand, the agentic dimension is task-oriented and defined by
independent, masterful, and assertive tendencies (e.g., competitive, ambitious,
dominating), with men scoring higher on this dimension than women (Eagly, 1987;
Eagly & Johannesen-Schmidt, 2001; Spence & Buckner, 2000).
Some literature suggests that the stereotypically female characteristic of
nurturance is valued less in the labor market than the stereotypically male characteristic
of aggressiveness (Kilbourne & England, 1996). Furthermore, women who have
internalized traditional beliefs about how women should act may be less likely to
behave assertively when they do not get promotion (Betz & Fitzgerald, 1987). Both
situations suggest that women with traditional gender role orientations would aspire
less to gain board seats, while those with no traditional role orientations would be more
inclined to gaining board appointments.
Social institutions and gender role attitudes
The fact that societies undertake socialization to enable men and women to fit
in their prescribed roles, suggests that the institutional environment plays a significant
8
part in defining gender role beliefs. This also suggests that there should be variation
from one context to the other or between countries in terms of gender role attitudes and
consequently women’s access to key positions of authority.
Several studies have examined how countries differ in their institutional
environments, leading to differences in practices and how such practices are interpreted
by their respective social institutions (Aguilera & Jackson, 2003; Kostova, 1999). For
example, Kostova (1999) provides a compelling explanation on how a country’s
institutional profile is configured. The country institutional profile refers to “the set of
all relevant institutions that have been established over time, operate in that country,
and get transmitted into organizations through individuals” (Kostova, 1997: 180). The
country institutional profile can be understood through three components namely,
regulatory (e.g., laws and rules) cognitive (e.g., shared knowledge), and normative (e.g.,
attitudes, values and norms). The three components form three pillars of Scott’s (1995)
conception of social institutions.
As mentioned earlier, societies develop gender role attitudes, and people
conform to such expectations through the way they behave both at home and work.
Gender role attitudes are, therefore, a function of social institutions that uphold and
transmit norms and values. This suggests that if social institutions change, gender role
attitudes may follow suit. We argue, therefore, that countries with social institutions
that place or facilitate less emphasis on traditional gender roles are likely to witness
equality or near equality of opportunities by both men and women and vice versa.
Indeed, research suggests that the salience of a particular stereotype may be dictated by
context (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987; van Knippenberg, van
Twuyver, & Pepels, 1994). For instance, the stereotypical belief that women do or do
not possess the qualities to be board directors may depend on the institutional
9
environment. In the context of this study such beliefs may vary from one country to the
next.
In this paper, we consider three important social institutions that have the
potential to differentially shape gender role attitudes across countries, namely the
political system (representation of women in parliament), the economic system and the
degree of religiosity. Of the many social institutions, the gender literature suggests that
the three in our study are arguably the most important factors of the institutional
environment for developing the taken-for-granted elements of gender roles (Lindsey,
2005). Moreover, while not tested empirically, Jackson’s (1998) thesis is that the
evolution of economic and political systems explains the levels of gender inequality
across countries. In the sections below, we discuss the link between these key social
institutions and the presence of women on boards, providing our hypotheses in the
process.
Representation of women in politics
We argue that it is plausible that gender stereotypes and their attendant effects
could be observed at varying levels in different institutions within a given period of
time. Specifically, political science literature shows that progress by women in politics
and government was made earlier (Skjeie, 1991) and far exceeds that made in business.
For example, in Norway the proportion of women in the Norwegian government and
politics has since been higher, compared to other areas of the labor market such as
academia and the private sector (Seierstad & Healy, 2012; Seierstad & Opsahl, 2011).
However, such other areas, in particular, the private sector are catching up. We argue
that this development is partly due to the fact that the successful entry of women in
politics and government encourages other women who are in business to aim higher
too, putting themselves forward for board seats.
10
This observation illustrates two important issues. First, it lends support to the
central argument in this paper that social institutions vary in their potential to break
down traditional gender social roles. Second, it shows that the breaking down of gender
role stereotypes in one part of society could serve as a reference point to initiate social
change in others. In this case, the breaking down of gender role stereotypes is catalyzed
by social learning, a process by which an individual learns from his neighbors’
experiences or social referents about their previous decisions and outcomes, behavior
or practices (Munshi, 2003). Here, the importance and motivation of learning from
social referents, via observation, imitation, and modelling cannot be underestimated
(Bandura, 1977). This suggests that much of human behavior is learned through the
influence of example. In the context of this study, seeing more women in government
or politics should increase the tendency of the observer (women in business) to aspire
for equivalent roles in business. Consequently, this effectively reduces the traditional
gender role attitudes by both men and women in the corporate sector.
There are similarities between the two groups (in politics and business) of
women. For instance, like board directors, members of parliament or government
ministers are elected or appointed on the basis of their knowledge, skills and experience.
Moreover, there have been calls for both business and politics to appoint women to the
highest echelons of leadership (Seierstad & Healy, 2012).
With such similarities in the backgrounds for the two groups of women and in
the requisite qualifications for the tasks, it would be expected that success achieved by
one part is likely to be an encouragement for the other. We argue, therefore, that
countries that have achieved success i.e. higher proportions of female members in high
level politics have managed to reverse traditional gender role attitudes in one area and
are more likely to repeat this in others, including on corporate boards.
11
We, therefore, suggest that the extent to which women are represented in
politics may determine the proportion of female directors in other key institutions,
including on corporate boards.
Hypothesis 1. The level of women’s representation in politics will positively impact on
the prevalence of women on boards of corporate directors.
Economic freedom
Earlier, we argued that gender stereotypes are likely to be broken down through
social change, whereby occupational and domestic-work segregation weaken, thus
changing the perception of men and women attributes. We argue that economic
development has the potential to provide this social change. Indeed, economic growth
increases the number of job vacancies that need to be filled, at all levels.
Two things are likely to happen that would mean more women in employment
and potentially in high-ranking posts. The first is that an expanding economy suggests
that with time there would be a shortage of skilled labor from the traditional sources.
Thus, where companies had been recruiting more men for executive posts, they may
find difficulties to carry on doing this as the competition intensifies. Companies would
then seek to recruit from alternative sources, hence the appointment of women across
all levels of the company hierarchy. Secondly, because of the high demand for labor
that results from a growing economy, companies may offer higher wages thus
encouraging more women to take paid jobs as opposed to staying home. Both scenarios
carry significant and positive changes in the context of gender role attitudes and
subsequent appointment to the upper echelons of the organization. Indeed, increased
participation in the workforce by women leads to the adoption of new gender role
attitudes as more women work alongside men.
12
We argue, therefore, that countries that are more economically developed are
more likely to have less traditional gender role attitudes, suggesting that more women
would be appointed to corporate boards. As Matland (1998: 114) notes: “Development
leads to weakening of traditional values, decreased fertility rates, increased
urbanization, greater educational and labor force participation for women, and
attitudinal changes in perceptions of the appropriate roles for women.” In the context
of our study, we therefore expect that women are more likely to be appointed as
directors on boards in a country that has developed economic institutions. We,
therefore, hypothesize:
Hypothesis 2. The institutional development of a country’s economic environment is
positively impacts on the prevalence of women on boards of corporate directors.
Religiosity
Religions, the shared set of beliefs, activities, and institutions based on faith in
supernatural forces (Stark & Bainbridge, 1985), are important social institutions in most
societies (Iannaccone, 1991). Consequently, religious doctrine has been known to
characterize variation in country attitudes in general, thus may, in particular, play an
important role in shaping attitudes toward women’s place in leadership.
With reference to the longstanding issue of gender inequality, scholars agree on
the continued role of religion in maintaining the status quo (Bendroth, 1993;
Woodhead, 2006). In fact, religion’s central role in consolidating gender differences
and inequality was recognized, explored and critiqued by nineteenth-century feminists
like Elizabeth Cady Stanton (Woodhead, 2006). In recent years, studies have found that
levels of gender equality across different countries are related to cultural factors – above
all, religiosity (Inglehart & Norris, 2003). Such studies show, too, that cross-sectional
differences in support for gender equality vary even between societies at similar levels
13
of development, and depend upon the degree of religiosity. This has prompted Inglehart
& Norris (2003) to conclude that, ‘religion matters, not only for cultural attitudes but
for the opportunities and constraints on women’s lives, such as the ratio of females to
males in educational enrolment, the female adult literacy rate, the use of contraception
as well as for opportunities for women in the paid workforce’ (2003: 69). Religion,
therefore, matters for opportunities for women to gain board appointments. This is not,
however, a question of religious men simply imposing religious attitudes upon women,
for traditional gender values and roles tend to be shared by both sexes in the same type
of society (Woodhead, 2007).
In general, religious institutions often prescribe what is considered right and
wrong, acceptable and unacceptable (Turner, 1997) and tend to be conservative or
patriarchal in their views about the place of women, both in the church hierarchy and
in society (Kenworthy & Malami, 1999). We add that 1religions can and do promote
gender stereotypes by their portrayal of ideal relations between the sexes (e.g.,
veneration of Mary, mother of Jesus; no women in Catholic priesthood). Indeed,
previous research suggests that most religions encourage and reinforce values
consistent with traditional gender roles (Lindsey, 2005; Sjoberg, 2004) “that include
separate subordinate positions for women” (Banaszak & Plutzer, 1993: 149).
Given the arguments we make above, therefore, most religions tend to promote
distinct gender roles, and consistent with social role theory most societal members use
religion to justify such arrangements. Therefore, countries with higher levels of
religiosity are more likely to have individuals exposed to views promoting traditional
gender roles (Lindsey, 2005). From such societies, it is more likely that firms, through
1
We are grateful to one of our reviewers for this suggestion.
14
their agents (CEOs, Chairmen and nomination committees), view traditional gender
roles as the norm, suggesting fewer board appointments for women.
Following these arguments we propose that:
Hypothesis 3. The degree of religiosity in a society relates negatively to the
proportion of female directors on the board.
Methodology
Sample and data
Our study explores the social triggers and barriers that explain the prevalence
of women on corporate boards across the world. To achieve our objectives, we use a
sample of 45 countries, taking an average of observations over seven years from 2007
through 2013. These countries are drawn from all the continents, however, in varying
proportions. Table 1 shows the distribution of the countries by region. The highest
number of countries is from Europe (16), followed by Asia (13). Table 1 below shows
the sample of countries in six regions.
Insert Table 1 about here
Measures
Dependent variable
Our dependent variable is the average percentage of women on boards
(womboard) for the years 2007 through 2013 in a given country. Data on this variable
was collected from the annual reports prepared by GMI. For example, the GMI report
15
titled GMI Ratings’ 2013 Women on Boards Survey contains data for the 45 countries
for the years 2008 through 2013. Data for 2007 was collected from surveys carried out
by Catalyst. We triangulated this data by using McKinsey & Company Report 1
abovefor the year 2007. Since 2007, McKinsey has been researching the business case
for increasing the number of women in senior management roles, published in the series
titled Women Matter. While not all the countries in our sample are covered in this
report, we used the ones provided therein to verify the reliability of the data provided
by Catalyst.
Independent variables
Our approach involves generating a set of independent variables that capture
key social institutions namely economic, political and religion. For Hypothesis 1 we
use the average percentage of women in parliament (womparly) for the years 2007
through 2013. We collect data for this variable from the World Bank Development
Indicators, published on their website.
For Hypothesis 2 we use the logarithm of the average gross domestic product
(lngdp) per capita as the basis of our measurement for level of economic development.
We also collect data for this variable from the World Bank Development Indicators,
published on their website.
For Hypothesis 3, we use the religiosity index to proxy for a country’s beliefs
and cultural values. The religiosity index represents the percentage of the population
who self-describe themselves as a ‘religious person’ in the question worded as:
Irrespective of whether you attend a place of worship or not, would you say you are a
religious person, not a religious person or a convinced atheist? Data on religiosity is
sourced from WIN-Gallup International and from the World Religion Database
(WRD). The former is the principal source while the latter was used to triangulate the
16
data. The WIN-Gallup International ‘Religiosity and Atheism Index’ which measures
global self-perceptions on beliefs is based on interviews with more than 50, 000 men
and women selected from countries across the globe (WIN-Gallup International,
2012).The World Religion Database contains detailed statistics on religious affiliation
for every country of the world. It provides source material, including censuses and
surveys, as well as estimates for every religion, offering a definitive picture of
international religious demography.
Control variables
We control for a specific cultural dimension with a direct link to gender roles
namely masculinity (Hofstede, 1988). According to Hofstede, (2001: 297),
“masculinity stands for a society in which social gender roles are clearly distinct: men
are supposed to be assertive, tough, and focused on material success; women are
supposed to be more modest, tender, and concerned with the quality of life.” Consistent
with social role theoretical arguments, societies with high scores of masculinity are less
likely to have more female directors.
In our second regression model, we also control for region to capture regional
characteristics. We use broad regional classifications as follows Europe, Eastern
Europe, Middle East and Africa, South America, North America and Asia. With 6
regions, we model 5 dummy variables. Region is considered an extremely broad
ideological variable (Paxton, 1997), so ideally, we would explain any impact of
regionalism with the other variables included in the model. If region has a direct impact
on the proportion of women on boards after the other variables are included, then we
assume that would be an indication of some cultural variation not explained within the
model.
Model specification
17
Using STATA we estimate the following regression:
𝑊𝑜𝑚𝑏𝑜𝑎𝑟𝑑𝑖 = 𝛽0 + 𝛽1 𝑤𝑜𝑚𝑝𝑎𝑟𝑙𝑦𝑖 + 𝛽2 𝑙𝑛𝐺𝐷𝑃𝑖 + 𝛽3 𝑟𝑒𝑙𝑖𝑔𝑖 + 𝛽4 𝑚𝑎𝑠𝑐𝑢𝑙𝑖 +
𝑒𝑖
[Eq. 1]
where women on board (womboard) is the dependent variable, women in parliament
(womparly), the logarithm of GDP per capita (lngdp) and religiosity are endogenous
regressors and masculinity (mascul) is a control variable that is taken as exogenous.
As pointed out above, our independent variables are potentially endogenous, that is they
may be systematically related to unobserved causes of the dependent variable, in this
case, the proportion of women on boards. Violations of this sort commonly occur when
factors related to independent variables, that predict outcome, are omitted from the
regression model or when the independent variables themselves are measured with error
(Woodridge, 2002). Of course, two-way causation is yet another condition but not the
only concern.
We thus used the 2SLS estimator or the instrumental variable (IV) method, an
econometric method which is useful to purge coefficients of endogeneity bias (Baum,
Schaffer, & Stillman, 2010) due to common methods, measurement error or
simultaneity (Antonakis, et al., 2010).
To use an IV method, the first step is to identify a set of variables that are
exogenous and that do not depend on other variables or disturbances in the system of
equations (Antonakis, Bendahan, Jacquart & Lalive, 2010). They should also be
sufficiently strongly related to the problematic endogenous covariates (Bollen,
2012).Moreover, the relationship between IVs and explanatory variables must be
justified by economic theory, passing the ‘theoretical overidentification’ test before an
18
empirical one (Antonakis, et al., 2010: 1104). Indeed, scholars (Antonakis, et al., 2010;
Reeb, Sakakibara & Mahmood, 2012) contend that identifying suitable instruments that
satisfy the conditions of relevance and exogeneity is the biggest challenge in applying
the IV method.
Once suitable instruments have been identified, the next step is to estimate the
coefficients in the regression model using two-stage least squares (2SLS) or similar
estimation methods. In this study we use 2SLS to estimate the coefficients in the
regression model shown above. 2SLS allows for consistent estimation of simultaneous
equations where one or more predictors are endogenous. This is because it handles
endogeneity resulting from omitted variables, measurement error, simultaneity, and
common method bias (Antonakis et al., 2010; Cameron & Trivedi, 2005; Greene, 2008)
Instrumental variables
We identified instrumental variables for the three explanatory variables. For the
explanatory variable-women in parliament, we use four instrumental variables namely
proportional representation, political quotas, ethnic and linguistic fractionalization
and number of years since first woman was elected to parliament.
We provide the rationale for each of these instruments. First, political scientists
accept the hypothesis that the presence of a proportional representation system
promotes female office-holding (Rule, 1987), since the system correlates the number
of seats won by a political party with the number of votes cast for their party. In
proportional representation systems, citizens vote for a party’s list rather than individual
candidates, suggesting that women need not stand alone to achieve political office
(Paxton, 1997), hence no bias attributable to gender. Second, gender political quotas
have the guarantee that a certain proportion of women will be appointed members of
parliament (Tripp & Kang, 2008). We, therefore, expect both proportional
19
representation and political quotas to have a positive impact on the number of female
members of parliament.
Proportional representation equals one for each year in which candidates were
elected using a proportional representation system; equals zero otherwise; averaged
for the years 1975-2000. This historical data, taken from Beck, Clarke, Groff, Keefer,
&Walsh (2001), cannot be changed by voting choices of the period 2007-2013, the
sample period of this study.
Fractionalization measures are based on the probability that two randomly
drawn individuals (from a country) are not from the same group (ethnic) or do not speak
the same language (linguistic). Both ethnic (ethnicfrac) and linguistic (linguifrac)
fractionalization variables are drawn from Alesina, Devleeschauer, Easterly, Kurlat &
Wacziarg (2003), and refer to the situation in the late 1980s and early 1990s. Scholars
argue that highly fragmented countries are more likely to affect the composition and
strength of political institutions such as parliament and government (Boubakri, Mansi
& Safar, 2013). We argue that countries that are highly fractionalized, both from an
ethnic and linguistic points of view, have over the years learnt to accept and to handle
diversity. Such countries would, therefore, be more accepting of yet another form of
diversity: gender. We, therefore, expect a positive association between ethnic and
linguistic fractionalization on one side and the proportion of female members of
parliament on the other. Having been collected in the 1980s -1990s and the fact that the
fragmentation of societies on the basis of ethnicity and language cannot be changed by
the dynamics of board gender parity in recent years, suggest that these instruments are
exogenous.
An additional instrument for women in parliament is the number of years since
the first women was elected to parliament (firstFemMP). We argue, therefore, that the
20
presence of females in parliament is determined by institutional path dependencies
(Terjesen, Aguilera & Lorenz, 2014), indeed a function of history. Path dependence
describes the ‘causal relevance of preceding stages in a temporal sequence’ Pierson,
2000: 252). Once a certain path has been chosen, future decisions are significantly
influenced by a previous one, and more so in an incremental manner (Greener, 2005).
Thus, countries with a longer history of appointing women to parliament are more likely
to have more female members of parliament in our sample period. As a historical fact,
defined by a specific time period, this variable cannot be influenced by the proportion
of women in parliament and is free from any possible shock in the disturbances, and
thus makes a good instrument.
For the explanatory variable, lnGDP, we use an exogenous geographical feature
of countries (latitude) as an instrument. Latitude, the distance from the equator
measured in absolute degrees, affects climate and, therefore, aspects of agricultural
productivity and disease. Previous studies have shown that countries closer to the
equator are less developed. Latitude has, therefore, considerable explanatory power for
the log of per capita GDP (Barro &McCleary, 2003), suggesting a positive relationship
between latitude and economic growth. Latitude is completely exogenous and can be
used as an appropriate instrument as cannot be changed by economic growth nor with
any level of shock from a disturbance.
We also instrument economic growth with legal origin. We use a set of dummy
variables for British, French, German, Scandinavian or Socialist legal origins. La Porta
et al (1998) show that whether a country’s commercial or company law is based on
British, French, German, Scandinavian or Socialist legal origins is important for
explaining the country’s laws on creditors’ rights, shareholder rights and private
property rights as well as the country’s level of bank and stock market development.
21
Furthermore, Levine (1999) traces the effect of legal origin on financial development
through to long-run economic growth, suggesting that legal origin influences economic
growth by shaping national financial systems. The basic thrust of the legal origin theory
is that common law (British), as opposed to French civil law and (to a lesser extent)
German and Scandinavian civil law, is associated with more orientation towards
institutions of the market instead of state intervention. According to studies based on
this theory, common law countries tend to be economically more developed. Legal
origin is a historical and static variable, reflecting the old age legal practices and thus
today’s economic growth does not determine a country’s legal origin.
For the explanatory variable-religiosity, we use two instrumental variables
namely state regulation of religion and religious pluralism. An important theory of
religiosity is the religion-market model (Barro & McCleary, 2003). According to this
theory, religiosity depends on government intervention, including the regulation of the
religion market by the state. Such an approach reduces the creativity and freedom of
religious beliefs to spread and thus should have a negative impact on religiosity. State
regulation (value 1 and 0 otherwise) refers to a situation in which the state appoints or
approves church leaders. This designation comes from discussions in Barrett, (1982)
and Barrett, Kurian and Johnson (2001) and typically applies in the late 1970s. Data on
state regulation is taken from Barro and McCleary (2003).
Religious pluralism, measured by an index of the diversity of religions that exist
in a country, is an attitude or policy regarding the diversity of religious belief systems
co-existing in society. Early scholarly work based on the secularization hypothesis
(Hume, 1757; Berger, 1967) provides the view that having more religious worldviews
makes them common thus more human and less divine (Berger, 1967). While this may
increase religious participation (the number of people attending churches, temples,
22
synagogues, etc) however, it may lead to less religiosity (the number of people who are
religious irrespective of attending a place of worship or not) in society. We, therefore,
expect religious pluralism, as an instrumental variable, to have a negative impact on
religiosity. Data on religious pluralism is taken from Barro and McCleary (2003). This
data, like that on state regulation is in many ways historical, applying in the late 1970s,
and thus makes both instrumental variables exogenous.
Results
The means, standard deviations and correlations among the study variables are
presented in Table 2. The average for the proportion of women on boards across the
45 countries in our sample is 9.24%, while that of women in parliament is higher at
22.77%. The average for religiosity is 59.68%. Correlations among the variables are
generally modest.
Insert Table 2 about here
As explained above, we tested the hypotheses using ivregress of STATA to
obtain 2SLS estimates. The results of the tests are shown in Tables 3 and 4. As
suggested by several scholars (e.g., Antonakis et al., 2010; Larcker & Rusticus, 2010),
we report both the OLS and the 2SLS results. From Table 3, Model 1 shows OLS
results. The coefficients for women in parliament (β =.25, p < .001) and religiosity (β
=-.07, p < .001) are as predicted. On average, a 1% increase in the number of women
in parliament would translate to a 0.25% increase in the number of female directors on
the board. On the other hand, a 1% increase in religiosity in a given country would lead
to a decrease in the number of female directors by 0.07%. However, from the OLS
results there is no support for the hypothesis that economic development is related to
women on boards.
23
Turning to the 2SLS estimation of the same equation, we start by explaining
tests on the suitability of the instruments by (1) making reference to the significance
of instrumental variables, in first-stage regressions, that predict endogenous variables,
(2) testing whether 2SLS is the preferred method to OLS by showing tests of
endogeneity (Hausman test), (3) using the score chi square (Sargan) statistic to validate
the exogeneity of instruments (test of over-identification) and (4) reporting the firststage F-statistic.
Models 2, 3, 4 and 5 report the 2SLS results. Models 2, 3, and 4 report first stage
estimations, where women in parliament, lngdp and religiosity are respectively taken
as endogenous regressors. Starting with the instruments for the variable, women in
parliament, the coefficients on proportional representation (β =7.61, p < .001), on the
period since the first women was elected to parliament (β =.13, p < .1) and on linguistic
fractionalization (β = 9.10, p < .1) are all positive and significant as expected.
The first instrument for economic development, latitude, is positive and
significant (β =.03, p < .001) as expected. As instruments, legal origins are significant
and positively related to economic development with varying coefficient sizes.
Specifically, British (β =2.21, p < .001), French (β =1.42, p < .001), German (β =1.48,
p < .001), and Scandinavian (β =1.61, p < .05)
As instruments, religious pluralism (β =-39.91, p < .05) and state regulation (β
=-19.32, p < .05) are both negative and significantly correlated to the dependent
variable religiosity as expected.
We also carried out Wu-Hausman test for each of the endogenous regressors
based on their respective instruments (results shown in Table 3 under the respective
variables). The endogenous regressor women in parliament (womparly) [χ2(1)=4.40,
24
p=.04] is endogenous, while
economic freedom (lngdp) [χ2(1)=.06, p=.81] and
religiosity [χ2(1)=1.58, p=.22] are both shown as exogenous.
While these tests for single endogenous regressors are indicative of the strengths
and relevance of the instruments used, joint tests including all variables and their
respective instruments tend to provide a better explanation for the state of variables and
for the suitability of their instruments. We, therefore, turn to results for the full model
or joint test, i.e. in the second stage of the 2SLS.
Contrasting the results of the second-stage regression with OLS estimation, the
Wu-Hausman test, for the full model, rejects the null that there is no endogeneity
problem with OLS estimation [χ2(3)=3.37, p<.05], suggesting that the 2SLS estimation
is preferred to the OLS estimation.
Using the Sargan test, we then examined the veracity (over-identification
restriction) of the 2SLS model to ensure that the excluded instruments from the equation
do not correlate with the disturbance of the dependent variable. The null hypothesis for
the Sargan test is that the instruments are exogenous, i.e. uncorrelated with the error
term. Thus, if the Sargan statistic is insignificant, we fail to reject the null hypothesis.
As shown in Model 5 the Sargan test for excluded instruments indicate that the model
constraints were tenable [χ2(9)=13.54, p=.14], suggesting that the instruments as a
group are exogenous. A post-estimation first-stage F-statistic test result of 22.46, p <
.001, suggests that the instruments are not weak.
Insert Table 3 about here
Compared to the OLS estimates, discussed above, coefficients from 2SLS trend
in the same direction, however, they are larger. For example, in Hypothesis 1, we
proposed that the proportion of women in parliament positively impacts on the
prevalence of women on boards of directors. As predicted, the coefficient for this
25
variable is significantly positive (β =.42, p < .001). On average, a 1% increase in the
proportion of women in parliament increases the number of female directors on the
board by 0.42%.
In Hypothesis 2, we proposed that economic development positively impacts on
the prevalence of female directors on boards. There is no support for this hypothesis as
the coefficient for economic development is not significant.
In Hypothesis 3, we proposed that religious beliefs negatively impact on the
prevalence of female directors on boards. As we predicted, religiosity has a negative
and significant impact on female board appointments (β =-0.12, p < 0.05). This
suggests that a 1% increase in religiosity in a given country would lead to a decrease in
the number of female directors by 0.12%.
Shea’s Adjusted R-square results for the independent variables are reasonably
high at 36%, 35% and 27% for women in parliament, economic freedom and religiosity
respectively.
We reran the tests, controlling for regional effects, as shown in the equation below.
𝑊𝑜𝑚𝑏𝑜𝑎𝑟𝑑𝑖 = 𝛽0 + 𝛽1 𝑤𝑜𝑚𝑝𝑎𝑟𝑙𝑦𝑖 + 𝛽2 𝑙𝑛𝐺𝐷𝑃𝑖 + 𝛽3 𝑟𝑒𝑙𝑖𝑔𝑖𝑜𝑠𝑖𝑡𝑦𝑖 + 𝛽4 𝑚𝑎𝑠𝑐𝑢𝑙𝑖 +
𝑅𝑒𝑔𝑖𝑜𝑛 + 𝑒𝑖
[Eq. 2]
Results from these tests are not significantly different from those reported in
Table 3. We thus carried out further post-estimation tests (testparm in Stata) to establish
the relevance of regional effects in our model. Results show that, for this study, regional
effects do not matter for both OLS [F(5, 35)=0.38, p > F=.86] and 2SLS [χ2(5)=2.74,
p > χ2=.74].
Discussion and conclusion
26
The starting point of this study is that people in a society observe the activities
of men and women and form corresponding beliefs about their psychological attributes.
From the different activities of the sexes, they infer gender stereotypes-shared
expectations that women and men are intrinsically different. Guided by gender role
beliefs that are shared within a society, citizens are socialized for the skills, traits, and
preferences that support their society’s division of labor. This, we argue, could
potentially explain the limited number of female directors on boards across societies.
However, some countries have higher proportions of female directors than others, a
situation we attribute to the variation of national social institutions. The study, thus,
examines the social and economic factors that promote or inhibit the appointment of
women to corporate boards of directors.
First, we find that the representation of women in parliament impacts positively
on the prevalence of women on boards. Drawing on social role theory, we argue that
this finding explains that the presence of female directors in other key societal
institutions helps to reduce traditional gender stereotypes, leading to acceptance of
women beyond the politics. The presence of women in parliament and, by extension,
in government could lead not only to acceptance of equal gender roles by both sexes,
but could also motivate women to aspire for higher positions in several other sectors
including in business. The implication for this finding is that there is need to improve
on the social institutions that promote the appointment of female directors to boards.
For example, mechanisms that help improve the proportion of female members of
parliament such as proportional representation and political quotas for female seats in
parliament have an indirect influence on the proportion of female directors, and yet
these, as a starting point, may be far easier to achieve. In that regard, more studies on
other areas that may have impact on board composition or on the determinants of the
27
same, such as on the proportion of women in the workforce or women in management,
are necessary.
The second hypothesis, that predicts a positive causal effect of economic
freedom on the proportion of female directors, is not significantly supported. This lack
of support could be explained by the fact that there are some developed countries with
low levels of female representation (e.g., Japan) and some with high levels of female
representation (e.g., Norway). The same could be said about countries with low levels
of development.
The third hypothesis, also drawing on social role theory, emphasizes that
religion encourages and reinforces values that are consistent with traditional gender
roles. Unlike other social institutions discussed in this paper, therefore, religiosity has
a negative effect on the prevalence of women on boards, and such was our hypothesis.
This hypothesis was supported. While it is easier to change the other social institutions,
changing attitudes that emanate from cultural institutions such as religion is difficult to
achieve. As such, understanding religious dynamics in society is a central task of studies
on gender, in general, and on women on boards, in particular. The long arm of the
Catholic church’s opposition to birth control, for example, demonstrates the power of
religion in enforcing traditional gender roles.
Several other messages, consistent with dominant cultural injunctions are
observable across religious groups. Moreover, religion adds to these injunctions the
message that they are natural and part of the universe, serving as an important vehicle
for the internalization of cultural codes and at the same time making them more rigid
and difficult to alter. In the context of this study, religion plays a key role in the
dynamics of women’s lack of representation on boards, and could potentially be used
to change this situation, however, difficult. It is, therefore, important not only to
28
understand the complexities of the relationship between religion and culture but to
change the way world religions work in order to achieve gender parity in corporate
boards and beyond.
Drawing on a key societal lens i.e. social role theory and recognizing the role
of social institutions in either upholding or discouraging gender stereotypes, the study
demonstrates the salience of the institutional environment in understanding board
gender diversity, offering a number of contributions. First, this paper contributes to the
growing corporate governance literature on female top executive leadership by
introducing theoretical perspectives from social psychology and gender studies to frame
the analysis – as called for by various researchers (e.g., Adams & Flynn, 2005; Terjesen
et al., 2009).
Second, drawing on corporate governance, religious, economic and gender
studies, this paper provides an explanation for the current state of board gender diversity
and insights to gender role attitudes in a relatively large number of countries. Moreover
and relatedly, a further contribution beyond the substantial findings is that this study is
one of the first to use a 2SLS procedure to identify institutional determinants of female
board appointments. To our knowledge, this statistical approach, that ensures
consistency of estimates in the face of endogeneity, has not been used in research on female
board appointments, reducing confidence in their causal claims. We, therefore, heed recent
calls by scholars (Antonakis et al., 2010; Larcker & Rusticus, 2010; Reeb et al., 2012)
who urge researchers to take endogeneity problems more seriously.
Third, this study provides enhanced understanding of the particular institutional
forces that impede or facilitate women's participation in corporate leadership in a
number of countries, including non-Western societies. Such knowledge assists policy
makers and development workers (e.g. from the UN) to craft programs and policies that
29
effectively address gender equality in general and women participation in corporate
affairs in particular. Moreover, such knowledge could help firms that operate in specific
countries to understand the dynamics of local environments on matters that concern
their practices on board composition and equal opportunities.
Indeed, this research has implications for organizational leaders. To start with,
the country variables investigated in this study are easily observable and measurable,
and, therefore, provide important insights for managers. For companies that decide to
operate in specific countries, setting up their headquarters there, we consider that
managers could react to institutional contexts that are conducive to non-acceptance of
gender diversity on boards. Based on our results they could decide to enter countries
with favorable institutions with greater possibility of utilizing both female and male
talent on their boards. Alternatively, they may enter countries with institutional barriers
and then utilize female talent as a competitive advantage, since on average many firms
will not be inclined to appoint female directors.
Moreover, while companies may not change the country’s institutions, they may
draw up internal procedures and practices that may improve the position of women not
only for the sake of board appointments but for motivating female employees in
general.
We also consider that our paper has implications for policy makers. For
example, the fact that political office by women encourages more female board
appointments suggests that governments may need to examine their political policies,
if they are to achieve greater female board presence. As such, governments with fewer
women members of parliament or in other public social spheres may come up with
policies that encourage or facilitate gender parity in such areas. Such efforts may
include legislating in favor of women empowerment, alongside targeted training
30
programs to ensure that women succeed in these roles. Moreover, since societies that
are highly religious are less likely to achieve board gender parity, we suggest that policy
makers should push for a mandatory approach to female board appointments, such as
the use of board quotas, indeed, a current argument in governance studies (Tejersen et
al., 2014).
Limitations and areas for future research
Notwithstanding the timeliness of this research and the contributions discussed
above, we acknowledge a few limitations of this study. First, while we use data that is
sufficiently rich, and from reliable sources, to test our hypotheses, secondary data have
some obvious disadvantages, as we did not have full control of sample selection or the
development of our dependent variable measure. Despite the use of a strong statistical
corrective technique in IV estimation, future studies could provide an alternative
approach based on interviews, surveys and case studies. Second, as is common in
empirical research, we cannot tell exactly whether our variables measure what we think
they do. For example, we did not consider the specific aspects of religiosity to be able
to determine the exact and fine-grained effect of religion on the presence of women on
corporate boards. Future studies could distinguish the types of religious affiliation and
their respective effect on female board appointments.
Third, at the center of the discussion between religion and gender, in general,
and female directors, in particular, is the issue of power. Both religion and gender are
centrally implicated in unequal distributions of power, and that their interplays serve
and seek to reinforce existing distributions of power or to change them – in various
ways and by various means. Future studies could explore power as a key intervening
variable between religiosity and the appointment of female leaders.
31
Moreover, the institutional context of our study is limited to only forty-five
countries. Although the sample has countries from every region of the world and the
issues considered are universal in nature (e.g., religion), undergirded by a social
psychological lens in social role theory, suggesting a reasonable level of the
generalizability of our findings, however, the regional representation is not evenly
distributed. As country data on board composition continues to get better in more
countries, further studies could benefit from larger samples.
While this paper focuses on issues of gender, many of the arguments made in it
may also apply to other minority groups with diverse backgrounds (e.g., race or
ethnicity, sexual orientation, disability, or age). As is well documented, members of
these groups are also disadvantaged due to their group’s low social status (e.g., LorenziCioldi, 2006) and this may extend to cases of lower representation in corporate
leadership (Catalyst, 2007). Moreover, in many national and organizational contexts,
these groups of minorities are also likely to be perceived as atypical leaders.
In conclusion, this paper has provided factors that explain the variation of
female board appointments across countries. Being part of the general theme of gender
equality and the attendant benefits of diversity in corporations, the presence of females
on boards, as a study area, continues to be of great importance. A central theme in this
paper is that the lack of female board members resides in differences in societal roles
that lead men and women to demonstrate and value different types of interpersonal
behavior. As such corporate leadership through the institution of the board is considered
as the preserve of men. This paper provides the thesis that as a social construct, these
gender role stereotypes can be better understood by examining the social institutions
that potentially uphold or minimize this perception. Such factors include the proportion
of women in government/parliament, economic development and religiosity. The study
32
finds support for the hypotheses that the presence of more female members in
parliament increases female board appointments and that more religious societies tend
to have fewer females on corporate boards.
Acknowledgements
We are grateful to the Associate Editor, Professor John Antonakis and three anonymous
reviewers for their invaluable comments and suggestions that led to significant
improvement of this paper. We are also grateful to Joao X, Regina Frank and Ellen
Chizema for a number of conversations on the theme of women empowerment and
indeed their noticeable absence in key leadership positions.
References
33
Adams, S. M., & Flynn, P.M. (2005). Local knowledge advances women's access to
corporate boards. Corporate Governance: An International Review, 13(6): 836-846.
Aguilera, R.V., & Jackson, G. (2003). The cross-national diversity of corporate governance:
Dimensions and determinants. Academy of Management Review, 28(3): 447-465.
Alesina, A., Devleeschauer, A., Easterly, W., Kurlat, S., & Wacziarg, R. (2003).
Fractionalization. Journal of Economic Growth, 8(2): 155-194.
Antonakis, J., Bendahan, S., Jacquart, P., & Lalive, R. (2010). On making causal claims: A
review and recommendations. The Leadership Quarterly, 21(6): 1086-1120.
Banaszak, L., & Plutzer, E. (1993). Contextual determinants of feminist attitudes: National
and subnational influences in Western Europe. American Political Science Review,
87: 147–157.
Bandura, A. (1977). Social learning theory. New York: General Learning Press.
Barrett, D.B. (1982). World christian encyclopedia, 1st ed. Oxford: Oxford University Press.
Barrett, D.B., Kurian, G.T., & Johnson, T.M. (2001). World christian encyclopedia, 2nd ed.
Oxford: Oxford University Press.
Barro, R.J., & McCleary, R.M. (2003). Religion and economic growth. American
Sociological Review, 68: 760-781.
Beck, T., Clarke, G., Groff, A., Keefer, P., &Walsh, P. (2001). New tools in comparative
political economy: The database of political institutions. World Bank Economic
Review, 15(1): 165-176.
Bendroth, M. L. (1993). Fundamentalism and gender: 1875 to the present. New Haven: Yale
University Press.
Berger, P.L. (1967). The sacred canopy: Elements of a sociological theory of religion.
Garden City NJ: Doubleday.
Betz, N.E., & Fitzgerald, L.F. (1987). The career psychology of women. New York:
Academic Press.
Bollen, K. A. (2012). Instrumental variables in sociology and the social sciences. Annual
Review of Sociology, 38(1): 37-72.
Boubakri, N., Mansi, S., & Saffar, W. (2013). Political Institutions, connectedness and
corporate risk taking. Journal of International Business Studies, 44: 195-215.
Browne, K. (1999). Divided labors: An evolutionary view of women at work. New Haven,
CA: Yale University Press.
34
Bullough, A., Kroeck, K.G., Newbury, W., Kundu, S.K., & Lowe, K.B. (2012). Women’s
political leadership participation around the world: An institutional analysis. The
Leadership Quarterly, 23: 398-411.
Cameron, A. C., & Trivedi, P. K. (2005). Microeconometrics: Methods and applications.
New York: Cambridge University Press.
Catalyst, (2007). Catalyst census of women board directors of the Fortune 500.
Carter, D. A., D'Souza, F., Simkins, B. J., & Simpson, W. G. (2010). The gender and ethnic
diversity of US boards and board committees and firm financial performance.
Corporate Governance: An International Review, 18(5): 396-414.
Conyon, M.J., & Mallin, C. (1997). Women in the boardroom: Evidence from large UK
companies. Corporate Governance: An International Review, 5: 112-117.
Eagly, A. H. (1987). Sex differences in social behavior: A social-role interpretation.
Hillsdale, NJ: Erlbaum.
Eagly, A. H. (2009). The his and hers of prosocial behavior: An examination of the social
psychology of gender. American Psychologist, 64(8): 644–658.
Eagly. A.H., & Johannesen-Schmidt, M.C. (2001). The leadership styles of women and men.
Journal of Social Issues, 57: 781-797.
Eagly, A.H., & Karau, S.J. (2002). Role congruity theory of prejudice toward female leaders.
Psychological Review, 109: 573-598.
Eagly, A.H., & Wood, W. (2012). Social role theory. In P. van Lange, A. Kruglanski, & E. T.
Higgins (Eds.), Handbook of theories in social psychology (Vol. 2, pp. 458-476).
Thousand Oaks, CA: Sage Publications.
Eyring, A., & Stead, B.A. (1998). Shattering the glass ceiling: Some successful corporate
practices. Journal of Business Ethics, 17: 245-251.
Fouts, H.N. (2008). Father involvement with young children among the Aki and Bofi
foragers. Cross-Cultural Research, 42: 290-312.
Glick, P., & Fiske, S.T. (2001). An ambivalent alliance: Hostile and benevolent sexism as
complementary justifications for gender inequality. American Psychologist, 56: 109118.
Goldberg, S. (1993). Why men rule: A theory of male dominance. Chicago: Open Court.
Greene, W. H. (2008). Econometric analysis (6th ed.). Upper Saddle River, NJ: Prentice–
Hall.
Greener, I. (2005). The potential of path dependence in political studies. Politics, 25: 62-72.
35
Greenhaus, J.H., & Parasuraman, S. (1999). Research on work, family, and gender: Current
status and future directions. In G. N. Powell (Ed.), Handbook of gender and work (pp.
391-412). Thousand Oaks, CA: Sage.
Grosvold, J., & Brammer, S. (2011). National institutional systems as antecedents of female
board representation: An empirical study. Corporate Governance: An International
Review, 19(2): 116–135.
Hall, P.A., & Soskice, D. (2001). Varieties of capitalism in the institutional foundations of
comparative advantage. Oxford: Oxford University Press
Harris, R., & Firestone, J. (1998). Changes in predictors of gender role ideologies among
women: A multivariate analysis. Sex Roles, 38:239–52.
Heilman, M.E. (2001). Description and prescription: How gender stereotypes prevent
women’s ascent up the organizational ladder. Journal of Social Issues, 57, 657-674.
Hillman, A. J., Shropshire, C., & Cannella, A.A. (2007). Organizational predictors of women
of corporate boards. Academy of Management Journal, 50: 941–952.
Hofstede, G. (1998). Religion, masculinity and sex. In G. Hofstede & Associates (Eds.),
Masculinity and femininity: The taboo dimension of national cultures (pp.192-209).
Thousand Oaks, CA: Sage.
Hofstede, G. (2001). Culture's consequences: Comparing values, behaviors, institutions and
organizations across nations. Thousand Oaks CA: Sage Publications.
Hume, D. (1757). The natural history of religion. Oxford: Oxford University Press.
Iannaccone, L.R. (1991). The consequences of religious market structures: Adam Smith and
the economics of religion. Rationality and Society, April: 156-177.
Ingelhart, R., & Norris, P. (2003). Rising tide. Gender equality and cultural
change
around the world. Cambridge: Cambridge University Press.
Jackson, R.M. (1998). Destined for equality: The inevitable rise of women’s status. Harvard:
Harvard University Press.
Jackson, G., & Deeg, R. (2008). Comparing capitalisms: Understanding institutional diversity
and its implications for international business. Journal of International Business
Studies, 39: 540-561.
Jost, J.T. (2006). The end of the end of ideology. American Psychologist, 61: 651-670.
Jost, J.T., & Kay, A.C. (2005). Exposure to benevolent sexism and complementary gender
stereotypes: Consequences for specific and diffuse forms of system justification.
Journal of Personality and Social Psychology, 88, 498-509.
36
Kenworthy, L., & Malami, M. (1999). Gender inequality in political representation: A
worldwide comparative analysis. Social Forces, 78(1): 235-269.
Kilbourne, B., & England, P. (1996). Occupational skill, gender, and earnings. In P. Dubeck
& K. Borman (Eds.) Women and work: A handbook. New York: Garland.
Kirchmeyer, C. (2002). Gender differences in managerial careers: Yesterday, today and
tomorrow. Journal of Business Ethics, 37: 5 – 24
Koenig, A. M., Eagly, A. H., Mitchell, A. A., & Ristikari, T. (2011). Are leader stereotypes
masculine? A meta-analysis of three research paradigms. Psychological Bulletin, 137:
616–642.
Kostova, T. (1997). Country institutional profile: Concepts and measurement. Best Paper
Proceedings of the Academy of Management: 180-184.
Kostova, T. (1999). Transnational transfer of strategic organizational practices: A conceptual
perspective. Academy of Management Review, 24: 208-324.
Larcker, D. F., & Rusticus, T. O. (2010). On the use of instrumental variables in accounting
research. Journal of Accounting and Economics, 49: 186-205.
Lindsey, L.L. (2005). Gender roles - A sociological perspective. Upper Saddle River, NJ:
Prentice Hall.
Lorenzi-Cioldi, F. (2006). Group status and differentiation. In T. Postmes & J. Jetten (Eds.),
Individuality and the group (94–115). Sage: London.
Matland, R.E. 1998. Women’s legislative representation in national legislatures: a
comparison of democracies in developed and developing countries. Legislative
Studies Quarterly, 28: 109–125.
McKinsey & Company, (2007). Gender diversity, a corporate performance driver, Women
Matter, 1-25.
Munshi, K. (2003). Social learning in a heterogeneous population: technology diffusion in the
Indian Green Revolution. Journal of Development Economics, 73: 185-213.
Nielsen, S., & Huse, M. (2010). The contribution of women on boards of directors: Going
beyond the surface. Corporate Governance: An International Review, 18(2): 136-148.
OECD, (2009). Gender and sustainable development: Maximizing the economic, social, and
environmental role of women. OECD: Paris.
Paxton, P. (1997). Women in national legislatures: A cross-national analysis. Social Science
Research, 26: 442-464.
Peterson, C., & Philpot, J. (2007). Women’s roles on U.S. Fortune 500 boards: Director
expertize and committee memberships. Journal of Business Ethics, 72: 177-196.
37
Pierson, P.(2000). Increasing returns, path dependence, and the study of politics. American
Political Science Review, 94: 251-267.
Reeb, D., Sakakibara, M., & Mahmood, I. P. (2012). From the Editors: Endogeneity in
international business research. Journal of International Business Studies, 43(3): 211218.
Richerson, P. J., & Boyd, R. (2005). Not by genes alone: How culture transformed human
evolution. Chicago: University of Chicago Press
Ridgeway, C. (2006). Linking social structure and interpersonal behavior: A theoretical
perspective on cultural schemas and social relations. Social Psychology Quarterly,
69:5-16.
Ruigrok, W., Peck, S., & Tacheva, S. (2007). Nationality and gender diversity on Swiss
corporate boards. Corporate Governance: An International Review, 15: 546-557.
Rule, W. (1987). Electoral systems, contextual factors and women’s opportunity for election
to parliament in twenty-three democracies. The Western Political Quarterly, 40: 477498.
Sayer, L., Cohen, P.N., & Casper, L. (2004). Women, men and work. In The American
People: Census 2000. Russell Sage Foundation & Population Reference Bureau.
Scott, R. (1995). Institutions and organizations. Thousand Oaks, CA: Sage Publications.
Seierstad C., & Opsahl, T. (2011). For the few not the many? The effects of affirmative
action on presence, prominence, and social capital of female directors in Norway.
Scandinavian Journal of Management, 27 (1): 44-54.
Seierstad, C., & Healy, G. (2012). Women’s equality in the Scandinavian academy–a distant
dream? Work, Employment and Society, 26(2): 296-313.
Skjeie, H. (1991). The rhetoric of difference: On women’s inclusion into political elites.
Politics and Society, 19(2): 233–63.
Singh, V., & Vinnicombe, S. (2004). Why so few women in top UK boardrooms? Evidence
and theoretical explanations. Corporate Governance: An International Review, 12:
479-488.
Spence, J. T., & Buckner, C. E. (2000). Instrumental and expressive traits, trait stereotypes,
and sexist attitudes. Psychology of Women Quarterly, 24: 44–62.
Stark, R., & Bainbridge, W.S. (1987). A theory of religion. New York: Lang.
Stock, J.H., & Yogo, M. (2002). Testing for weak instruments in linear IV regression. In
D.W.K. Andrews & J.H. Stock (Eds.), Indentification and infererence for econometric
38
models: Essays in honor of Thomas Rothenberg (pp.80-108). Cambridge: Cambridge
University Press.
Tennie, C., Call, J., & Tomasello, M. (2009). Ratcheting up the ratchet: on the evolution of
cumulative culture. Philosophical Transactions of the Royal Society of Biological
Sciences, 364: 2405-2415
Terjesen, S., Aguilera, R.V., & Lorenz, R. (2014). Legislating a woman’s seat on the board:
Institutional factors driving gender quotas for boards of directors. Journal of Business
Ethics, 50(2): 177-186.
Terjesen, S., & Singh, V. (2008). Female presence on corporate boards: A multi-country
study of environmental context. Journal of Business Ethics, 83(1): 55–64.
Terjesen, S., Sealy, R., & Singh, V. (2009). Women directors on corporate boards: A review
and research agenda. Corporate Governance: An International Review, 17: 320-333.
Tripp, A.M., & Kang, A. (2008). The global impact of quotas: On the fast tract to increased
female legislative representation. Comparative Political Studies, 41(3): 338-361.
Turner, J.H. 1997. The institutional order. New York: Addison Wesley Educational
Publishers.
Turner, J.C., Hogg, M.A., Oakes, P.J., Reicher, S.D., & Wetherell, M.S. (1987).
Rediscovering the social group: A self- categorization theory. Oxford: Blackwell.
van Knippenberg, A., van Twuyver, M., & Pepels, J. (1994). Factors affecting social
categorization processes in memory. British Journal of Social Psychology, 33: 419431.
VanYperen, N.W., & Buunk, B.P. (1991). Equity theory, exchange orientation and communal
orientation from a cross-national perspective. Journal of Social Psychology, 131: 520.
WIN-Gallup International, (2012). Religiosity and Atheism Index. 1-25.
Woodhead, L. (2006). Sex and secularization. In G. Loughlin (ed.), Queer theology:
Rethinking the western body (pp. 230-234). Oxford: Blackwell.
World Bank. (2007). Doing business report: Comparing regulations in 178 countries.
Washington D.C.: The World Bank.
Todd, M., Johnson, M., & Grim, B. J. (eds.) (2003). World Religious Database.
39
Table 1: Distribution of countries by region
Region
Number of
Countries
13
16
4
4
3
5
Asia
Europe
Eastern Europe
South America
North America
Africa & Middle East
Total
45
Table 2: Means, standard deviations and correlations
Mean SD
1
2
1.womboard
9.24
2.womparly
22.77 10.69
0.60**
3. lngdp
9.73
0.39**
0.50**
4. relig
59.68 26.04
-0.39**
-0.29
3
6.38
1.13
40
-0.50**
4
5. mascul
51.07 19.26
-0.43*
-0.39
-0.11
-0.03
Notes: **p < .05; *p < 0.1
Abbreviated variables: women on the board (womboard), women in parliament (womparly), logarithm of gross
domestic product (lngdp), religiosity (relig), masculinity (mascul)
Table 3: OLS and 2SLS regression results: Dependent variable- proportion of women on the board
Shea’s
Adj R2
Model 1
OLS
Model 2
(1st stage)
womparly
Model 3
(1st stage)
lngdp
Model 4
(1st stage)
religiosity
Model 5
2nd Stage
2SLS
Intercept
14.40**
(9.78)
9.67
(9.34)
6.84***
(.96)
124.62***
(25.55)
26.74**
(13.53)
womparly
.26***
(.09)
.42***
(.12)
.36
lngdp
-.20
(.88)
-1.67
(1.20)
.35
relig
-.07**
(.03)
-.12**
(.05)
.27
41
-.09**
(.04)
-.04
(.08)
-.00
(.01)
-.24
(.22)
polquota
-4.71
(3.37)
.87**
(.35)
9.92
(9.21)
proprep
7.61***
(2.55)
.21
(.26)
-15.28**
(6.96)
firstFemMP
.13*
(.07)
-.00
(.01)
-04
(.19)
ethnifrac
.67
(7.15)
.31
(.73)
30.98
(19.56)
linguifrac
9.10*
(5.93)
-1.74***
(.61)
.38
(16.23)
latitude
.13
(0.10)
.03***
(.01)
-.76***
(.27)
Leg_Brit
5.47
(4.87)
2.21***
(.50)
-14.45
(13.32)
Leg_Frenc
8.61*
(5.08)
1.42***
(.52)
-7.37
(13.91)
Leg_Germ
4.79
(4.89)
1.48***
(.50)
-14.15
(13.36)
Leg_Scan
16.26**
(7.33)
1.61**
(.75)
-20.44
(20.05)
relplural
16.39**
(7.30)
1.78**
(.75)
-38.91**
(19.98)
stateregrel
.79
(2.97)
-.15
(.30)
-19.32**
(8.13)
.72
.74
.65
χ2(1)=
χ2(1)=
χ2(1)=1.58,
mascul
R2
.48
-.07
(.04)
.41
χ2(9)= 13.54, p=.14
Sargan
Hausman
4.40, p=.04
.06, p=.81
p=.22
F-Statistic
F(12, 31)= 5.37***
F(12, 31)= 7.29***
F(12,31)= 4.85***
Notes: Standard errors in parentheses, N= 45, *p<0.10, ** p<0.05, *** p<0.001
χ2(3)= 3.37, p=.03
22.46***
Abbreviated variables: women in parliament (womparly), logarithm of gross domestic product (lngdp), religiosity (relig),
masculinity (mascul), political quotas (polquota), proportional representation (proprep), number of years since a female member
was appointed to parliament (firstFemMP), ethnic fractionalization (enthnifrac), linguistic fractionalization (linguifrac), legal
British origin , (Leg_Brit), legal French origin (Leg_Frenc), legal German origin (Leg_Germ), legal Scandinavian origin
(Leg_Scan), religious pluralism (relplural), state regulation of religion (stateregrel).
42